Abstract:Cloud cover can significantly hinder the use of remote sensing images for Earth observation, prompting urgent advancements in cloud removal technology. Recently, deep learning strategies have shown strong potential in restoring cloud-obscured areas. These methods utilize convolution to extract intricate local features and attention mechanisms to gather long-range information, improving the overall comprehension of the scene. However, a common drawback of these approaches is that the resulting images often suffer from blurriness, artifacts, and inconsistencies. This is partly because attention mechanisms apply weights to all features based on generalized similarity scores, which can inadvertently introduce noise and irrelevant details from cloud-covered areas. To overcome this limitation and better capture relevant distant context, we introduce a novel approach named Attentive Contextual Attention (AC-Attention). This method enhances conventional attention mechanisms by dynamically learning data-driven attentive selection scores, enabling it to filter out noise and irrelevant features effectively. By integrating the AC-Attention module into the DSen2-CR cloud removal framework, we significantly improve the model's ability to capture essential distant information, leading to more effective cloud removal. Our extensive evaluation of various datasets shows that our method outperforms existing ones regarding image reconstruction quality. Additionally, we conducted ablation studies by integrating AC-Attention into multiple existing methods and widely used network architectures. These studies demonstrate the effectiveness and adaptability of AC-Attention and reveal its ability to focus on relevant features, thereby improving the overall performance of the networks. The code is available at \url{<a class="link-external link-https" href="https://github.com/huangwenwenlili/ACA-CRNet" rel="external noopener nofollow">this https URL</a>}.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the impact of cloud cover on earth observation in remote - sensing images. Specifically, clouds block the earth's surface, resulting in approximately 67% of remote - sensing data being unusable, which seriously affects further analysis and applications. Therefore, there is an urgent need for advanced cloud removal techniques to restore the areas blocked by clouds. Although the existing cloud removal methods are effective to a certain extent, they have some limitations: 1. **Traditional methods**: These include space - based, spectral - based, time - based, and hybrid methods. These methods usually rely on additional data (such as spectral, time, multi - sensor inputs), which increases the complexity of restoration and performs poorly when dealing with cloud cover in complex scenes. 2. **Deep - learning methods**: In particular, methods combining convolutional neural networks (CNN) with attention mechanisms show great potential in restoring areas blocked by clouds. However, these methods often lead to image blurring, artifacts, and inconsistencies because the attention mechanism weights all features according to general similarity scores, inadvertently introducing noise and irrelevant details. To solve these problems, the author proposes a new attention mechanism - **Attentive Contextual Attention (AC - Attention)**. AC - Attention can effectively filter out noise and irrelevant features by dynamically learning data - driven attention selection scores, thereby better capturing relevant long - distance contextual information. By integrating the AC - Attention module into the DSen2 - CR cloud removal framework, the model's ability to capture important long - distance information is significantly improved, thus removing clouds more effectively. The main contributions of the paper include: 1. Introducing Attentive Contextual Attention (AC - Attention), which aims to capture information - rich long - distance features and eliminate irrelevant or unimportant patterns. 2. Integrating AC - Attention into ACA - CRNet to improve its cloud removal ability and demonstrating its effectiveness. 3. Evaluating this method on three different datasets. The results show that it is superior to existing cloud removal methods in both visual quality and quantitative indicators. In addition, the ablation study confirms the effectiveness and adaptability of AC - Attention in capturing important long - distance dependencies.

Attentive Contextual Attention for Cloud Removal

Attention Mechanism-Based Generative Adversarial Networks for Cloud Removal in Landsat Images

AFFNet: Attention Mechanism Network Based on Fusion Feature for Image Cloud Removal

Multi-Stage Frequency Attention Network for Progressive Optical Remote Sensing Cloud Removal

An Innovative Approach for Effective Removal of Thin Clouds in Optical Images Using Convolutional Matting Model

Thick Cloud Removal Under Land Cover Changes Using Multisource Satellite Imagery and a Spatiotemporal Attention Network

LGE-CR: Integrating Nonlocal Gated Attention and Two-Flow Residual Dense Learning for Cloud Removal

Diffusion Enhancement for Cloud Removal in Ultra-Resolution Remote Sensing Imagery

PMAA: A Progressive Multi-scale Attention Autoencoder Model for High-performance Cloud Removal from Multi-temporal Satellite Imagery

Denoising Diffusion Probabilistic Feature-Based Network for Cloud Removal in Sentinel-2 Imagery

Gated Convolutional Networks for Cloud Removal From Bi-Temporal Remote Sensing Images

Remote sensing image cloud removal based on multi-scale spatial information perception

DiffCR: A Fast Conditional Diffusion Framework for Cloud Removal From Optical Satellite Images

RICE: A Dataset and Baseline for Cloud Removal in Remote Sensing Images

Multi-Head Linear Attention Generative Adversarial Network for Thin Cloud Removal

Cascaded Memory Network for Optical Remote Sensing Imagery Cloud Removal

A Novel Spectral Indices-Driven Spectral-Spatial-Context Attention Network for Automatic Cloud Detection

A deep-learning reconstruction method for remote sensing images with large thick cloud cover